GC41D-0595 Low Flows over the Eastern United States (1962...

GC41D-0595 Low Flows over the Eastern United States (1962-2011): Variability, Trends, and AttributionsJonghun Kam1(jkam@princeton.edu) and Justin Sheffield1

1Department of Civil and Environmental Engineering, Princeton University, Princeton, NJ, 08648

AbstractWe examined trends and variability in low flows over the eastern U.S. (S. Carolina to Maine) and theirattribution in a changing climate. We selected 149 out of 4878 USGS stations over the eastern U.S., takinginto account data availability and minimal direct management. Annual 7-day low flows (Q7) were computedfrom the series of daily stream flow records for 1962-2011 and compared to an antecedent precipitation (AP)index calculated over the corresponding basin for each station.

A north-south increasing-decreasing pattern in low flow trends was found that is associated with trends inAP. The exception is in the southern part of the domain where increasing trends in AP may have beenoffset by water withdrawals and increasing potential evapotranspiration (PET) as driven by increasingtemperature and vapor pressure deficit. Teleconnections between detrended Q7 and nine atmospheric andoceanic climate indices indicate that the North Atlantic Oscillation (NAO) and Pacific North America (PNA)pattern show statistically significant correlations for Q7 at one and two month lead time, respectively. Ourfindings suggest that the risk of future hydrological droughts is increased during a strong negative phase ofthe NAO and positive phase of the PNA during the summer, and may be further enhanced with temperaturedriven increases in PET.

Conclusions1) A dipole pattern of increasing and decreasing (north-south) trends in Q7 low flows exists.

2) Decreasing trends in Q7 over the Mid-ATL region and the southern part of the SE regionincluding North and South Carolina and Virginia.

3) The decreasing trends are possibly linked to increasing trends in PET driven by warmingtemperature.

4) NAO (-) and PNA (+) : A favorable condition for drought over the Mid-ATL and SE regions.

Background: Low Flow Hydrology

Stream flow is a hydrological response of theland surface at the catchment scale toprecipitation and is a major water resourceavailable for human use and ecological needs.

During the dry season, low precipitation (lowsupply) and high evapotranspiration (highdemand) conspire to reduce stream flow andwater availability. This natural seasonalhydrological phenomenon is known as the lowflow period (Smakhtin, 2001). With variationsin climate, decreases in seasonal precipitationand increases in potential evapotranspirationcan lead to reductions in low flows, withadverse effects on human activities andecosystem function (Bradford and Heinonen,2008).

Drivers of low flow variability are complex, andinclude antecedent precipitation, atmosphericdemand, surface water management(urbanization, dams, reservoirs, andirrigation), and groundwater withdrawals.

Figure 1: (a) Time series of daily stream flow andprecipitation at one station (USGS ID: 01531500) for 1998.(b) Time series of annual 7-day low flows and 30-day totalantecedent precipitation at one to three month lead beforethe date of the annual low flow event (e.g. the antecedentprecipitation accumulated during 0-29, 30-59, and 60-89days, respectively) for 1962-2011.

Changes in Stream Flows over Eastern United States

• 1950s-1990s : Increasing (+) trends in low and median flows due to increasing fall precipitation •

• 2000s-2100s : Decreasing (-) trends in low flows (e.g. low flow volumes) due to increasing trends in temperature-driven potential evapotranspiration

Figure 2: Map of thelocations of the 149USGS HCDN stations andthe masks of theircorresponding upstreambasins. The sizes of bluecircles represent thesizes of the basins.

Attribution of Trends in Low Flows I: Antecedent Precipitation

• Northeastern Region: Increasing (+) trends in Q7 and AP-90

• Mid-Atlantic (Mid-ATL) Region: Decreasing (-) trends in Q7 and AP-90

Figure 3: Spatial distribution of the trends in annual 7-day low flows (Q7) and 90-day accumulatedantecedent precipitation (AP-90) from three versions of Mann Kendall test (MK-0, MK-1, MK-2) over1962-2011. Dots represent stations with no regulation.

Recent Weakening in Associations of Antecedent Precipitation

• Mid-ATL region: Weaker consensus of the trends in Q7 and AP-90 during 1982-2011 than during 1962-1991.

Figure 4: Same as Figure 3 except for the period ((a) and (c): 1962-1991 and (b) and (d): 1981-2011).

Attribution of Trends in Low Flow II: Potential Evapotranspiration

Atlantic Coastal Region:

• 1962-2011: Increasing (+) trends with 0.006 mm/day/year

• 1982-2011: Increasing (+) trends with 0.013 mm/day/year

Figure 5: Spatial distribution of the trends in six-month warm season (May-October) potentialevapotranspiration (PET) from the MK-0 test for (a) 1962-2011, (b) 1962-1991, and (c) 1982-2011.

Associations with Atmospheric Circulation

• North Atlantic Oscillation (NAO)

July-September (JAS) : Significant and positive (+) correlation with the Mid-Atlantic region

• Pacific North American (PNA) Oscillation

June-August (JJA): Significant and negative (-) correlation with West Virginia and Virginia

Figure 6: Map of the 3-month time period (colors) for each station which has the maximum temporalcorrelation between detrended Q7 and the (a) North Atlantic Oscillation and (b) Pacific-North Americanpattern, for 1962-2011.

Relevance of Results

Adverse effects for water and energy:

• Food Production

• Domestic Uses

• Public Health

• Electricity Production

• Ecological Quality

References1) Smakhtin VU (2001) Low flow hydrology: a review, J.

Hydrol., 240, 147–186.2) Bradford MJ, Heinonen JS (2008) Low Flows, instream

flow needs and fish ecology in small streams, Can. Water Resour. J., 33, 165–180.

AcknowledgementsThis work was supported by the USGS (G11AP20215)and NOAA (NA11OAR4310097). The authorsacknowledge access to the TIGRESS high performancecomputer center at Princeton University.